Hi Martin,
Commenting on your first post, in rhinoscript, you could enable the detail, after checking that the view is Rhino.IsLayout, then use Rhino.ViewNearCorners to get the array of exact corner 3d points of the open detail, then with this array of points use Rhino.ObjectsByRegion to select the objects in the modelspace, (option to select by window or crossing window).
In your 2nd post, perhaps you could then export (while the detail is still enabled) these selected objects to a file, re-enable the layout, copy the border and relevant detail to the clipboard.
Then open the new file created from ExportSelected using code, in the new file, create a layout to the correct size and paste the border and detail from the clipboard, then close the file again from code… or something along these lines. The border, detail and objects will all be the same locations as the original.
I think this would all be done as a Rhino file, then a further operation with code to save as .dwg, so you’d end up with 2 sets of files, which may (or may not) be beneficial for any further editing.